In the past, I’ve been regularly evangelizing on deep linking (like in my guest post on Apptamin or my talk at App Promotion Summit - slides), explaining the mechanics implied. I used to advize publishers to build it in-house, like we did at Deezer.
Not anymore - here’s why:

• Maturity: Existing solutions were quite clunky a few years ago, they since have matured and now provide much better integration process and reliability.
• TTM / Time To Market: Building it in-house = time before it’s ready. You want it now.
• Tracking: Setting up proper tracking / analytics on in-house built deep linking is just a nightmare. Really.
• Maintenance and changes: That’s the most important point. You’ll have to maintain the thing. Additionally, due to its platform and device dependencies, deep linking is really hard to monitor and requires (regular) manual testing. And then, it changes. You can’t imagine the mess it has been for some actors when iOS 9 changed everything / Apple universal links were introduced. Everything broke! So, you’ll also have to test regularly / over time, including when a new OS is launched. 
• Fragmentation & multiple formats: There are (or there can be) different behaviors throughout Android vendors, OS versions, … You also have to tackle with multiple formats, for Facebook app links or Twitter cards, so that your contents with deep links get optimized for rich media display.
• Innovation: Some really cool innovations appeared, based on deep linking. The most notable one is deferred deep linking. It uses the same mechanics than tracking for ads, and allows instead to personalized onboarding or advanced referral programs (in-app). That’s not something you can build in-house (requires R&D). 

Like many topics in tech, deep linking became complicated enough for dedicated solutions to be more and more relevant. Building it in-house and having it work properly in all cases requires a lot of work, maybe even dedicated resources. In fact, I estimate that to have a basic deep linking solution built in-house up & ready, it requires one third of the working time of an IT guy, maybe more depending the needs.
Then, as a publisher of apps, the wise question is: is it your core job to maintain a deep linking thing? Or should you rather work on your app and make it a great product?
Just as it’s not anymore relevant to host your own servers at home, route your emails yourself, implement push notifications in-house, … it is now most of the time not relevant anymore to build deep linking in-house.
​

I’ve mentioned many deep linking solutions in the past, but they have now either matured or disapeared / became irrelevant.
It’s a topic that attribution / campaigns analytics actors have seized well. Tune, Adjust or AppsFlyers / OneLink all provide their solution.
But there are also pure deep linking specialists, that bring really cool innovations like personalized onboarding: Yozio, Hokolinks, and Branch
NB: I’m currently really found of Branch, I’ve recommended them to many contacts and clients and they provided great satisfaction (btw, they already fuel deep linking for major publishers in EU: Dubsmash, Dojo, CDiscount, QuizzUp, …).
 

But there is also a whole new subject, strongly tied to deep linking: App Indexing, which I covered quite early at SMX Paris 2015 with @Polo_Seo (see  http://www.journaldunet.com/solutions/seo-referencement/app-indexing-au-smx-paris-2015.shtml in french), while it was only available to a few selected publishers for testing.

​App indexing is the mechanism that allows the contents of your app to be indexed into search engines. Be it Google (see https://developers.google.com/app-indexing/) or iOS Spotlight Search (tech documentation).
And deep linking is a requirement to app indexing, because it’s how contents get accessed from within search results.
 
I formerly hilighted deep linking as your « Everything2App », now we can add Search2App:

To know everything about App Indexing, you can read my guest post on Branch.io's blog: https://blog.branch.io/the-past-present-and-future-of-app-indexing

 
These are the slides of my talk at App Promotion Summit in London.
​The video on Business of Apps is viewable here (less than 15mns).
In addition to amazing articles and resources spotted during my strategic watch, I give many examples of apps I've been working on and for which we applied these techniques with success (Deezer, Viadeo, LeKiosk, Consult, ...), sometimes leading to millions of new users. Indeed, I estimate that Deezer and Spotify grew their apps downloads to tens of millions (and also retention) just with deep linking, viral loops and proper content SEO (at that time in 2010, we didn't have a marketing budget at all - startup mode!). 

First, my talk started with a brief introduction to startup life and to AARRR (Acquisition - Activation - Retention - Revenue - Referral), which is the typical funnel to focus on when building a startup. And that's also true for an app!
I prefer to use the cycle "Love, Money and Growth". Indeed, before even thinking about growth, you have to ensure your users love your product. On this matter, the concept of Minimum Lovable Product is quite interesting! (@AndrewChen's take on Minimum Desirable Product is great too)

So, as mentioned, "Love" for your product has to come first. It's indeed quite useless to try to grow a product people don't love (or don't love YET - indeed it's a search and discovery process, with iterations and pivots, and it's all about Product Management). During that phase, we aim to reach a status called Product / Market fit.
Then it's time to focus on growth, and to leverage growth hacking methodologies and techniques.
To get what growth hacking is, the article "Growth Hacker is the new VP Marketing" is definitely a must read (@AndrewChen again!)


In the slides, 7 growth hacks for mobile apps are detailed:

• Hack your existing marketing chanels: Deep linking is more and more an essential factor to leverage all your existing marketing assets, and is the ground to quite all other growth hacks for mobile apps. There are now many deep linking solutions providers: Tune, AppsFlyers, Adjust, URX, Deeplink.me, AppUrl, Tapstream... Some actors like Branch even use deferred deep linking to do quite advanced magic on personalization (more on that later / at the end of this article). Everything about deep linking here: http://www.apptamin.com/blog/mobile-deep-linking/

• Hack expectation: hack your landing page, and launch much prior to app to collect leads / Beta testers contacts. They'll help you build a better app and they'll be the 1st evangelizers of the app at launch.

• Hack SEO with App content: description of the app on appstores and landing page is quite poor. Why not index all the great content that is in the app? (SEO, App indexing, ...)

• Hack ASO: ASO relies most on optimizing the number of conversions to app install, compared to the number of people that visited your appstores pages. So, it's obvious that visual elements people can see on this formatted page are critical. The app name, texts, app icon and screenshots have to be crafted very well to ensure maximum impact - and optimizing these can lead to conversions increase of up to 30%. But how to test them? How to find what works, since there's not much ways to "try" on the appstores pages? Comes the "App store proxy" technique: we create a fake landing mobile web landing page that looks exactly like app store page, and we do all the tests on it. Then we replicate the findings on the app store pages. And there are tools for that: SplitMetrics, StoreMaven and TestNest

• Hack ratings and reviews: I won't provide any detail here, it has to be seen in the slides and this article has to be read - The Right Way To Ask User To Review Your App. Basically, this is a virtuous mechanism that favors good ratings & reviews, and helps you to collect feedback to build a better product and satisfy your users even more. Note that it's not new, the main evangelizers of this technique were Apptentive and Helpshift - both provide tools to make it work with your helpdesk.

• Hack virality: When WoM (Word of Mouth) works, social networks and sharing mechanics are just technical ways to amplify it. But it's then essential to growth and speed of growth. However, it's not something that can be created artificially, and it requires a lot of time (tests / measure / tweek) before getting real results / sustainable growth. In case of mobile apps, there are now nice tools to favor sharings (Facebook App Invites, Google App Invites, ...). Two channels are often forgotten when talking about sharing: email and SMS. And it appears these 2 ones are just powerful for virality, because they allow to share a specific content privately and to targeted contacts.

As quotted by Peggy Anne Salz:

• Hack referral: A referral program right within the app can be a huge growth driver, particularly if it's incentivized. And now there are amazing tools to create them easily with deep linking but also personalized on-boarding which dramatically increase the conversion. Indeed, without such technique (known as deferred deep linking), the app store acts as a "cycle breaker" when the app gets installed by the invitee. Deferred deep linking uses ad tracking mechanisms, but instead of triggering an ad attribution, these services trigger personalization on first app opening. I'm personally a huge fan of Branch (also known as BranchMetrics) both for their product and contents, but Yozio and Hokolinks are great tools too.

Last part of the slides showcases the app Consult (an app for doctors), where we mixed a lot of these techniques, including a nice way to ask your users to share your app. 

Finally, since I'm a product guy, I also have a slide deck on how to build a great app: 
http://fr.slideshare.net/AlexandreJubien/mobile-product-build-great-apps-producttank-paris-17


Feel free to reach out to me if you have any feedback to make the slides or the article better!

Slides from my talk "Build great mobile apps!" at ProductTank Paris #17. 
Many mobile product tips inside!